Soil moisture observation quality control method and system based on multi-source data fusion

Abstract

The invention provides a soil moisture observation quality control method and system based on multi-source data fusion. The method comprises the following steps: acquiring hourly soil volumetric water content data of an automatic soil moisture observation station and mode data of a soil volumetric water content analysis product of a land surface data assimilation system; performing data cleaning on the soil volumetric water content data to obtain effective observation data; interpolating the soil volumetric water content analysis product to the site position of the automatic soil moisture observation station by adopting a nearest neighbor interpolation method so as to form a space-time matching sequence of observation data and mode data; performing consistency check on the space-time matching sequence to obtain abnormal data; performing systematic deviation detection on the abnormal data to analyze the deviation change problem of the abnormal data to obtain deviation data; and correcting the deviation data, and performing interpolation filling on the corrected data to obtain a complete soil moisture data set. According to the invention, the quality and consistency of data can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of soil moisture monitoring, and in particular to a soil moisture observation quality control method and system based on multi-source data fusion. Background Art

[0002] Soil moisture is an important indicator for studying land-atmosphere interaction, soil erosion, crop moisture monitoring and agricultural meteorological forecasting. Understanding the changing patterns of soil moisture is of great significance to agricultural production and climate forecasting. However, the existing automatic soil moisture observation system is easily affected by factors such as equipment aging, sensor calibration parameter drift, and environmental interference during long-term operation, resulting in systematic deviations or anomalies in the observed data.

[0003] At present, some studies have proposed some soil moisture quality control methods, such as anomaly detection algorithms based on the relationship between ground temperature and precipitation, and analysis methods based on the change characteristics of soil hydrophysical constants. However, these methods still have shortcomings in detecting systematic deviations, especially the consistency analysis of observed data and background field data. In addition, traditional methods have limited supplementary capabilities in the case of missing data, and it is difficult to meet the needs of real-time monitoring and high-precision analysis. Summary of the invention

[0004] In order to overcome the shortcomings of the prior art, the purpose of the present invention is to provide a soil moisture observation quality control method and system based on multi-source data fusion, which not only improves the quality and consistency of the data, but also provides important basic data and analysis tools for scientific research, decision support and sustainable development.

[0005] To achieve the above object, the present invention provides the following solutions:

[0006] A soil moisture observation quality control method based on multi-source data fusion, including:

[0007] Obtain hourly soil volumetric moisture data from automatic soil moisture observation stations and model data of soil volumetric moisture analysis products from the Land Surface Data Assimilation System;

[0008] Performing data cleaning on the soil volume water content data to obtain valid observation data;

[0009] Using the nearest neighbor interpolation method, the soil volumetric moisture content analysis product is interpolated to the site location of the automatic soil moisture observation station to form a spatiotemporal matching sequence of the observation data and the model data;

[0010] Performing consistency check on the spatiotemporal matching sequence to obtain abnormal data;

[0011] Conduct systematic deviation detection on abnormal data to analyze the deviation change problem of abnormal data and obtain deviation data;

[0012] The deviation data is corrected, and the corrected data is interpolated and filled to obtain a complete soil moisture data set.

[0013] Preferably, the soil volumetric moisture content data covers multiple soil layer depths; the soil layer depth ranges include 0-10 cm, 10-20 cm and 20-30 cm.

[0014] Preferably, the spatial resolution of the model data is 0.0625°×0.0625°, and the soil layer depth is 0-10 cm, 10-40 cm and 40-100 cm.

[0015] Preferably, data cleaning is performed on the soil volume moisture content data to obtain valid observation data, including:

[0016] The soil volume moisture content data is grouped according to a preset collection period to obtain a plurality of data groups;

[0017] Calculate the difference coefficient between the current data group and the previous data group in sequence;

[0018] Determine whether the value of the coefficient of difference is within a preset range;

[0019] If the value of the difference coefficient is not within the preset range, the corresponding data group is removed;

[0020] If the value of the difference coefficient is within a preset range, the corresponding data group is retained until all data groups are traversed to obtain the observed data.

[0021] Preferably, the difference coefficient calculation formula is:

[0022]

[0023] Among them, p X,Y is the coefficient of difference, cov(X,Y) represents the covariance between the current data set X and the previous data set Y, α X represents the mean of the current data set X, β Y Represents the mean of the previous data set Y.

[0024] Preferably, the soil volumetric moisture content analysis product is interpolated to the site location of the automatic soil moisture observation station using the nearest neighbor interpolation method to form a spatiotemporal matching sequence of the observation data and the pattern data, including:

[0025] Extract the longitude and latitude coordinates of all automatic soil moisture observation stations to form a list of station locations;

[0026] For each observation site in the site location list, the distance between the observation site and the grid point of the model data is calculated; the distance d is calculated as: Among them, x 1 and x 2 are the latitude and longitude coordinates of the observation site and the latitude and longitude coordinates of the grid points of the model data, respectively; 1 and 2 are the latitude and longitude coordinates of the observation site and the latitude and longitude coordinates of the grid points of the model data, respectively;

[0027] Determine the nearest grid point, and use the soil volume moisture value of the grid point as the interpolation result;

[0028] The timestamps of the observation data and the pattern data are controlled to be consistent. If the time resolutions are different, the pattern data is time interpolated to match the time resolution of the observation data, so as to obtain the time-space matching sequence after time and space matching.

[0029] Preferably, the indicators for consistency check of the spatiotemporal matching sequence include: mean square error, root mean square error, mean absolute error and correlation coefficient; wherein the calculation formula of the correlation coefficient is: Among them, O i and M i are the observed value and the model value, respectively. and are the means of the observed values ​​and the mode values, respectively, and n is the number of samples.

[0030] Preferably, the deviation variation problem includes sensor calibration parameter drift, equipment performance degradation and equipment failure.

[0031] Preferably, the method of interpolating and filling the corrected data is Kriging interpolation technology.

[0032] A soil moisture observation quality control system based on multi-source data fusion, including:

[0033] A data acquisition module is used to obtain hourly soil volumetric moisture content data from the automatic soil moisture observation station and model data of soil volumetric moisture content analysis products from the land surface data assimilation system;

[0034] A data cleaning module, used for cleaning the soil volume moisture content data to obtain valid observation data;

[0035] A sequence generation module, used to interpolate the soil volumetric moisture content analysis product to the site location of the automatic soil moisture observation station using the nearest neighbor interpolation method, so as to form a spatiotemporal matching sequence of the observation data and the pattern data;

[0036] A consistency check module, used to perform consistency check on the spatiotemporal matching sequence to obtain abnormal data;

[0037] The deviation detection module is used to perform systematic deviation detection on abnormal data to analyze the deviation change problem of abnormal data and obtain deviation data;

[0038] The correction and filling module is used to correct the deviation data and interpolate and fill the corrected data to obtain a complete soil moisture data set.

[0039] According to the specific embodiments provided by the present invention, the present invention discloses the following technical effects:

[0040] The present invention provides a soil moisture observation quality control method and system based on multi-source data fusion, the method comprising: obtaining hourly soil volumetric moisture data of an automatic soil moisture observation station and pattern data of a soil volumetric moisture analysis product of a land surface data assimilation system; performing data cleaning on the soil volumetric moisture data to obtain valid observation data; using the nearest neighbor interpolation method to interpolate the soil volumetric moisture analysis product to the site location of the automatic soil moisture observation station to form a spatiotemporal matching sequence of the observation data and the pattern data; performing consistency check on the spatiotemporal matching sequence to obtain abnormal data; performing systematic deviation detection on the abnormal data to analyze the deviation change problem of the abnormal data to obtain deviation data; correcting the deviation data, and interpolating and filling the corrected data to obtain a complete soil moisture data set. The present invention not only improves the quality and consistency of data, but also provides important basic data and analysis tools for scientific research, decision support and sustainable development. The implementation of the present invention will help to better understand and manage soil moisture resources and promote progress in related fields. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0042] Figure 1 A flow chart of a method provided by an embodiment of the present invention;

[0043] Figure 2 A schematic diagram of the system structure provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0044] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0045] The purpose of the present invention is to provide a soil moisture observation quality control method and system based on multi-source data fusion, which improves the quality and consistency of data and provides important basic data and analysis tools for scientific research, decision support and sustainable development.

[0046] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0047] Figure 1 A flow chart of a method provided by an embodiment of the present invention, such as Figure 1 As shown, the present invention provides a soil moisture observation quality control method based on multi-source data fusion, comprising:

[0048] Step 100: Acquire hourly soil volumetric water content data of the automatic soil moisture observation station and model data of soil volumetric water content analysis products of the land surface data assimilation system;

[0049] Step 200: performing data cleaning on the soil volume water content data to obtain valid observation data;

[0050] Step 300: using the nearest neighbor interpolation method, interpolating the soil volumetric moisture content analysis product to the site location of the automatic soil moisture observation station to form a spatiotemporal matching sequence of the observation data and the model data;

[0051] Step 400: Perform consistency check on the spatiotemporal matching sequence to obtain abnormal data;

[0052] Step 500: Performing systematic deviation detection on abnormal data to analyze the deviation change problem of abnormal data and obtain deviation data;

[0053] Step 600: Correct the deviation data and interpolate and fill the corrected data to obtain a complete soil moisture data set.

[0054] Preferably, the soil volumetric moisture content data covers multiple soil layer depths; the soil layer depth ranges include 0-10 cm, 10-20 cm and 20-30 cm.

[0055] Preferably, the spatial resolution of the model data is 0.0625°×0.0625°, and the soil layer depth is 0-10 cm, 10-40 cm and 40-100 cm.

[0056] Specifically, step 100 of this embodiment includes:

[0057] 1. First, it is necessary to establish a connection with the automatic soil moisture observation station and the land surface data assimilation system to obtain the required data regularly. The automatic soil moisture observation station is usually equipped with sensors that can monitor the soil volumetric moisture content in real time and store the data in a local database. Automatic extraction of these data can be achieved through API or data transmission protocols (such as HTTP, FTP, etc.). Preferably, the acquired soil volumetric moisture content data should cover multiple soil layer depths, including 0-10cm, 10-20cm and 20-30cm, so as to fully understand the vertical distribution of soil moisture.

[0058] 2. Secondly, the model data of soil volume moisture analysis products provided by the land surface data assimilation system also need to be obtained. These model data are usually generated by numerical weather prediction models or soil moisture models, with high spatial resolution (such as 0.0625°×0.0625°) and cover different soil depths (such as 0-10cm, 10-40cm and 40-100cm). These model data can be downloaded regularly by accessing the database of the assimilation system or using the relevant API. Ensure that the downloaded data is aligned in time with the observation data for subsequent spatiotemporal matching analysis.

[0059] 3. Finally, the acquired observation data and pattern data are integrated and stored in a unified database. Relational databases (such as MySQL, PostgreSQL) or non-relational databases (such as MongoDB) can be used to store these data. When integrating data, it is necessary to ensure that the data format is consistent, including information such as timestamp, soil depth, and spatial coordinates. Through data cleaning and preprocessing, invalid data and outliers are removed to ensure the accuracy of subsequent analysis. The integrated data set will provide a basis for subsequent spatiotemporal matching, quality control, and analysis.

[0060] Through the above steps, this embodiment can effectively obtain and integrate the hourly soil volumetric moisture content data of the automatic soil moisture observation station and the model data of the land surface data assimilation system, laying a solid foundation for subsequent soil moisture monitoring and analysis.

[0061] Preferably, data cleaning is performed on the soil volume moisture content data to obtain valid observation data, including:

[0062] The soil volume moisture content data is grouped according to a preset collection period to obtain a plurality of data groups;

[0063] Calculate the difference coefficient between the current data group and the previous data group in sequence;

[0064] Determine whether the value of the coefficient of difference is within a preset range;

[0065] If the value of the difference coefficient is not within the preset range, the corresponding data group is removed;

[0066] If the value of the difference coefficient is within a preset range, the corresponding data group is retained until all data groups are traversed to obtain the observed data.

[0067] Preferably, the difference coefficient calculation formula is:

[0068]

[0069] Among them, p X,Y is the coefficient of difference, cov(X,Y) represents the covariance between the current data set X and the previous data set Y, α X represents the mean of the current data set X, β Y Represents the mean of the previous data set Y.

[0070] Since the data acquisition device may be affected by its own parameters or environmental factors, the measured value collected by the data acquisition device at a certain moment may deviate greatly from the actual value. Therefore, this application can screen out abnormal monitoring values ​​through the difference coefficient to ensure the accuracy of the data.

[0071] Preferably, the soil volumetric moisture content analysis product is interpolated to the site location of the automatic soil moisture observation station using the nearest neighbor interpolation method to form a spatiotemporal matching sequence of the observation data and the pattern data, including:

[0072] Extract the longitude and latitude coordinates of all automatic soil moisture observation stations to form a list of station locations;

[0073] For each observation site in the site location list, the distance between the observation site and the grid point of the model data is calculated; the distance d is calculated as: Among them, x 1 and x 2 are the latitude and longitude coordinates of the observation site and the latitude and longitude coordinates of the grid points of the model data, respectively; 1 and 2 are the latitude and longitude coordinates of the observation site and the latitude and longitude coordinates of the grid points of the model data, respectively;

[0074] Determine the nearest grid point, and use the soil volume moisture value of the grid point as the interpolation result;

[0075] The timestamps of the observation data and the pattern data are controlled to be consistent. If the time resolutions are different, the pattern data is time interpolated to match the time resolution of the observation data, so as to obtain the time-space matching sequence after time and space matching.

[0076] Optionally, step 300 of this embodiment includes:

[0077] 1. First, extract the latitude and longitude coordinates of all stations from the data of the automatic soil moisture observation station. These coordinates include the latitude and longitude information of each observation station to form a list of station locations. This list will serve as the basis for the subsequent interpolation process to ensure that the geographical location of each observation station can be accurately identified. Ensure the integrity and accuracy of the data to avoid errors in subsequent calculations.

[0078] 2. Next, for each observation site in the site location list, calculate the distance between it and the model data grid point. To do this, first obtain the longitude and latitude information of all model data grid points. By traversing each observation site and all grid points, calculate the distance between them to determine which grid point is closest to the observation site. This process usually involves simple geometric calculations to ensure that the closest grid point can be found effectively, thereby ensuring the accuracy of the interpolation result.

[0079] 3. After calculating the distances between all observation sites and grid points, determine the nearest grid point for each observation site. Use the soil volume moisture value of the grid point as the interpolation result and assign it to the corresponding observation site. This process ensures that the observation data can be effectively combined with the model data, so that the soil moisture information of the observation site can reflect the trend and changes of the model data.

[0080] 4. Finally, to ensure that the timestamps of the observed data and the model data are consistent, check the temporal resolution of the two. If the temporal resolution is different, the model data needs to be temporally interpolated to match the temporal resolution of the observed data. This can be achieved through linear interpolation or other appropriate temporal interpolation methods to ensure that a consistent spatiotemporal matching sequence can be formed in time. After completing this process, the final spatiotemporal matching sequence will provide a reliable data basis for subsequent soil moisture analysis and monitoring.

[0081] Preferably, the indicators for consistency check of the spatiotemporal matching sequence include: mean square error, root mean square error, mean absolute error and correlation coefficient; wherein the calculation formula of the correlation coefficient is: Among them, O i and M i are the observed value and the model value, respectively. and are the means of the observed values ​​and the mode values, respectively, and n is the number of samples.

[0082] Specifically, step 400 of this embodiment includes:

[0083] 1. First, extract observations and pattern values ​​from the spatiotemporal matching sequence. Based on these data, calculate multiple consistency test indicators, including mean square error, root mean square error, mean absolute error, and correlation coefficient. The mean square error and root mean square error are used to evaluate the overall deviation between the observations and the pattern values, while the mean absolute error provides an intuitive understanding of the error size. The correlation coefficient is used to measure the strength of the linear relationship between the observations and the pattern values. The calculation of these indicators will provide a basis for subsequent abnormal data identification.

[0084] 2. After calculating the consistency test indicators, you need to set reasonable thresholds for each indicator. These thresholds can be based on the statistical characteristics of historical data, the experience of field experts, or standards in relevant literature. By comparing the calculated indicator values ​​with the set thresholds, the consistency of the data can be judged. For example, if the root mean square error exceeds the set threshold, or the correlation coefficient is lower than a certain critical value, it may indicate the presence of abnormal data.

[0085] 3. Next, the data in the spatiotemporal matching sequence is marked according to the set threshold. Data points whose index values ​​exceed the threshold are marked as abnormal data. This process can be achieved by writing data processing scripts or using data analysis software to ensure that inconsistent data can be quickly and accurately identified. The marked abnormal data will be extracted separately for subsequent analysis and processing.

[0086] 4. Finally, the marked abnormal data are analyzed in depth to determine the causes and characteristics. This may include time series analysis, spatial distribution analysis, and comparison with other related variables. Through analysis, patterns and trends in abnormal data can be identified, which can provide a basis for subsequent data correction and quality control. This process not only helps to improve the reliability of data, but also provides more accurate information for soil moisture monitoring and management.

[0087] Preferably, the deviation variation problem includes sensor calibration parameter drift, equipment performance degradation and equipment failure.

[0088] Furthermore, step 500 of this embodiment includes:

[0089] 1. First, classify the abnormal data marked in step 400 and analyze its deviation characteristics. By comparing the deviation values ​​of the abnormal data and the pattern data, the type of deviation can be preliminarily determined. For example, if the deviation value shows a trend of continuous increase or decrease, it may be caused by the drift of the sensor calibration parameters; if the deviation value suddenly increases within a certain period of time, it may be caused by equipment performance degradation or failure. By analyzing the time series and spatial distribution of the deviation, a basis can be provided for subsequent deviation detection.

[0090] 2. Analyze the long-term trend of abnormal data for deviations that may be caused by drift in sensor calibration parameters. By fitting a time series curve of the deviation value (such as linear regression or polynomial fitting), determine whether there is a systematic change in the deviation. If the deviation value is found to gradually increase or decrease over time, it can be inferred that the sensor calibration parameters have drifted. Furthermore, this inference can be verified and the degree of drift can be quantified by combining historical data or referring to data from other observation sites.

[0091] 3. For deviation problems caused by equipment performance degradation, focus on analyzing the volatility and stability of the deviation value. By calculating the standard deviation or coefficient of variation of the deviation value, determine whether the equipment performance has abnormal fluctuations. If the deviation value shows great instability or randomness within a certain period of time, it may be caused by equipment performance degradation. In addition, the equipment performance problem can be further confirmed in combination with the equipment's operation records (such as maintenance logs or operation time).

[0092] 4. For deviation problems caused by equipment failure, focus on the sudden change of deviation values. By performing mutation point detection on the time series of deviation values ​​(such as sliding window method or CUSUM algorithm), identify significant changes in deviation values ​​at a certain moment. If the deviation value suddenly increases or decreases in a short period of time and the abnormality persists, it may be that the equipment has failed. Combined with the operating status of the equipment and the data from other observation sites, the existence of the fault can be further verified and provide a basis for the maintenance or replacement of the equipment.

[0093] Through the above steps, this embodiment can systematically detect the deviation change problem of abnormal data and attribute it to sensor calibration parameter drift, equipment performance degradation or equipment failure. This process provides a scientific basis for subsequent data correction and equipment maintenance, ensuring the quality and reliability of soil moisture observation data.

[0094] Preferably, the method of interpolating and filling the corrected data is Kriging interpolation technology.

[0095] Specifically, step 600 of this embodiment includes:

[0096] 1. First, select the appropriate correction method according to the characteristics of the deviation data. For the deviation caused by the drift of sensor calibration parameters, a linear regression or polynomial regression model can be used to fit the deviation, and the fitting result can be used to correct the data; for the random deviation caused by the degradation of equipment performance, the data can be smoothed by methods such as sliding average or Kalman filtering; for the sudden change deviation caused by equipment failure, it can be replaced or corrected by combining the model data or the observation data of the neighboring stations. The selection of the correction method should be based on the source and change characteristics of the deviation to ensure that the corrected data is as close to the true value as possible.

[0097] 2. After completing the deviation correction, the correction results need to be verified. The correction effect can be evaluated by calculating the consistency test indicators between the corrected data and the model data (such as mean square error, root mean square error, correlation coefficient, etc.). If the corrected data significantly improves the degree of match with the model data, it means that the correction method is effective; otherwise, the correction model or method needs to be readjusted. In addition, the rationality of the correction data can be further verified by comparing it with the observation data of neighboring stations.

[0098] 3. For missing values ​​or outliers that still exist in the corrected data, interpolation is required to generate a complete data set. Temporal interpolation (such as linear interpolation and spline interpolation) can be used to fill missing values ​​in the time series, or spatial interpolation (such as nearest neighbor interpolation and Kriging interpolation) can be used to fill data gaps in space. If there are many missing data, multi-source data fusion interpolation can be combined with model data or data from neighboring stations to improve the accuracy and reliability of the interpolation results.

[0099] 4. After completing the deviation correction and interpolation filling, integrate all processed data into a complete soil moisture data set. Ensure that the data set covers all observation sites, time points and soil depths, and the data format is unified and there are no missing values. The final data set can be stored as a structured file (such as CSV, NetCDF, etc.) and used for subsequent soil moisture analysis, model building or decision support. Through this process, this embodiment can effectively improve the integrity and quality of the data and provide a reliable data basis for soil moisture monitoring and management.

[0100] Corresponding to the above method, such as Figure 2 As shown, this embodiment also provides a soil moisture observation quality control system based on multi-source data fusion, including:

[0101] A data acquisition module is used to obtain hourly soil volumetric moisture content data from the automatic soil moisture observation station and model data of soil volumetric moisture content analysis products from the land surface data assimilation system;

[0102] A data cleaning module, used for cleaning the soil volume moisture content data to obtain valid observation data;

[0103] A sequence generation module, used to interpolate the soil volumetric moisture content analysis product to the site location of the automatic soil moisture observation station using the nearest neighbor interpolation method, so as to form a spatiotemporal matching sequence of the observation data and the pattern data;

[0104] A consistency check module, used to perform consistency check on the spatiotemporal matching sequence to obtain abnormal data;

[0105] The deviation detection module is used to perform systematic deviation detection on abnormal data to analyze the deviation change problem of abnormal data and obtain deviation data;

[0106] The correction and filling module is used to correct the deviation data and interpolate and fill the corrected data to obtain a complete soil moisture data set.

[0107] The beneficial effects of the present invention are as follows:

[0108] (1) The present invention cleans the soil volumetric moisture content data, removes outliers and noise, and ensures that the data used is valid and reliable. This helps to improve the accuracy of subsequent analysis and decision-making; the nearest neighbor interpolation method is used to match the model data with the observation data in time and space, which can effectively fill the data gaps in space and ensure the consistency of data from different sources in time and space. This is crucial for the dynamic monitoring of soil moisture.

[0109] (2) The present invention can identify abnormal data in a timely manner by performing consistency checks on the spatiotemporal matching sequence. This early identification mechanism helps to avoid the impact of erroneous data on subsequent analysis; performing systematic deviation detection on abnormal data can deeply analyze the source and change characteristics of abnormal data, providing a basis for data correction.

[0110] (3) The present invention corrects the deviation data, which can effectively correct the systematic errors between the observed data and the model data, and improve the accuracy and reliability of the data; by interpolating and filling the corrected data, a complete soil moisture data set is finally obtained. This integrity is crucial for subsequent soil moisture analysis, model building and decision support.

[0111] (4) The multi-source data fusion method of the present invention enables data from different sources to complement each other and provide more comprehensive soil moisture information. This is of great significance for scientific research and decision-making support in the fields of agricultural management, climate change research, ecological monitoring, etc.; through the hourly data collection of the automatic soil moisture observation station, combined with the analysis of the pattern data, it is possible to achieve real-time monitoring of soil moisture, providing timely decision-making basis for agricultural irrigation, soil management, etc.

[0112] (5) The cleaned, corrected and interpolated datasets of the present invention can significantly improve the prediction capability of the soil moisture model and enhance the model's ability to reflect actual conditions. Through accurate soil moisture monitoring and analysis, the use of water resources can be optimized, the sustainable development of agriculture can be promoted, water resource waste can be reduced, and crop yield and quality can be improved.

[0113] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the system disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0114] The principles and implementation methods of the present invention are described in this article using specific examples. The description of the above embodiments is only used to help understand the method and core idea of ​​the present invention. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting the present invention.

Claims

1. A soil moisture observation quality control method based on multi-source data fusion, characterized in that: include: Obtain hourly soil volumetric moisture data from automatic soil moisture observation stations and model data of soil volumetric moisture analysis products from the Land Surface Data Assimilation System; Performing data cleaning on the soil volume water content data to obtain valid observation data; Using the nearest neighbor interpolation method, the soil volumetric moisture content analysis product is interpolated to the site location of the automatic soil moisture observation station to form a spatiotemporal matching sequence of the observation data and the model data; Performing consistency check on the spatiotemporal matching sequence to obtain abnormal data; Conduct systematic deviation detection on abnormal data to analyze the deviation change problem of abnormal data and obtain deviation data; The deviation data is corrected, and the corrected data is interpolated and filled to obtain a complete soil moisture data set.

2. The soil moisture observation quality control method based on multi-source data fusion according to claim 1 is characterized in that: The soil volumetric moisture content data covers multiple soil layer depths; the soil layer depth ranges include 0-10 cm, 10-20 cm and 20-30 cm.

3. The soil moisture observation quality control method based on multi-source data fusion according to claim 1 is characterized in that: The spatial resolution of the model data is 0.0625°×0.0625°, and the soil layer depths are 0-10 cm, 10-40 cm, and 40-100 cm.

4. The soil moisture observation quality control method based on multi-source data fusion according to claim 1 is characterized in that: Data cleaning is performed on the soil volume moisture content data to obtain valid observation data, including: The soil volume moisture content data is grouped according to a preset collection period to obtain a plurality of data groups; Calculate the difference coefficient between the current data group and the previous data group in sequence; Determine whether the value of the coefficient of difference is within a preset range; If the value of the difference coefficient is not within the preset range, the corresponding data group is removed; If the value of the difference coefficient is within a preset range, the corresponding data group is retained until all data groups are traversed to obtain the observed data.

5. The soil moisture observation quality control method based on multi-source data fusion according to claim 4 is characterized in that: The coefficient of difference calculation formula is: Among them, p X,Y is the coefficient of difference, cov(X,Y) represents the covariance between the current data set X and the previous data set Y, α X represents the mean of the current data set X, β Y Represents the mean of the previous data set Y.

6. The soil moisture observation quality control method based on multi-source data fusion according to claim 1 is characterized in that: The soil volumetric moisture content analysis product is interpolated to the site location of the automatic soil moisture observation station using the nearest neighbor interpolation method to form a spatiotemporal matching sequence of the observation data and the model data, including: Extract the longitude and latitude coordinates of all automatic soil moisture observation stations to form a list of station locations; For each observation site in the site location list, the distance between the observation site and the grid point of the model data is calculated; the distance d is calculated as: Among them, x1 and x2 are the latitude and longitude abscissas of the observation site and the latitude and longitude abscissas of the grid points of the model data, respectively; y1 and y2 are the latitude and longitude ordinates of the observation site and the latitude and longitude ordinates of the grid points of the model data, respectively; Determine the nearest grid point, and use the soil volume moisture value of the grid point as the interpolation result; The timestamps of the observation data and the pattern data are controlled to be consistent. If the time resolutions are different, the pattern data is time interpolated to match the time resolution of the observation data, so as to obtain the time-space matching sequence after time and space matching.

7. The soil moisture observation quality control method based on multi-source data fusion according to claim 1 is characterized in that: The indicators for consistency check of the spatiotemporal matching sequence include: mean square error, root mean square error, mean absolute error and correlation coefficient; wherein the calculation formula of the correlation coefficient is: Among them, O i and M i are the observed value and the model value, respectively. and are the means of the observed values ​​and the mode values, respectively, and n is the number of samples.

8. The soil moisture observation quality control method based on multi-source data fusion according to claim 1 is characterized in that: The deviation change problems include sensor calibration parameter drift, equipment performance degradation and equipment failure.

9. The soil moisture observation quality control method based on multi-source data fusion according to claim 1 is characterized in that: The method of interpolating and filling the corrected data is the Kriging interpolation technique.

10. A soil moisture observation quality control system based on multi-source data fusion, characterized in that: include: A data acquisition module is used to obtain hourly soil volumetric moisture content data from the automatic soil moisture observation station and model data of soil volumetric moisture content analysis products from the land surface data assimilation system; A data cleaning module, used for cleaning the soil volume moisture content data to obtain valid observation data; A sequence generation module, used to interpolate the soil volumetric moisture content analysis product to the site location of the automatic soil moisture observation station using the nearest neighbor interpolation method, so as to form a spatiotemporal matching sequence of the observation data and the pattern data; A consistency check module, used to perform consistency check on the spatiotemporal matching sequence to obtain abnormal data; The deviation detection module is used to perform systematic deviation detection on abnormal data to analyze the deviation change problem of abnormal data and obtain deviation data; The correction and filling module is used to correct the deviation data and interpolate and fill the corrected data to obtain a complete soil moisture data set.

Images